Antibody-Targeted FISH Analysis Improves Detection of Residual Disease in “High Risk” B-Cell Acute Lymphoblastic Leukemia.

Abstract

A number of prognostically important genetic subtypes of acute lymphoblastic leukemia (ALL) has been identified. Most notably, the presence of t(9;22)/BCR-ABL1 or 11q23/MLL rearrangements in B-cell ALL is a poor prognostic indicator and patients with these subtypes of ALL are candidates for allogeneic stem cell transplantation. Consequently, early detection of minimal residual disease (MRD) is critical for appropriate diagnostic and therapeutic decisions. Current methods for MRD measurements are based on morphologic examinations, flow cytometry, quantitative reverse-transcription polymerase chain reaction (Q-RT-PCR) and cytogenetics/fluorescence in situ hybridization (FISH). While high abnormality rates ease disease detection through standard morphologic and cytogenetic analyses, the presence of cytogenetically-aberrant lymphoblasts at low levels post treatment hampers residual disease detection. In this study, we analyzed 126 samples collected from 51 patients using a sequential immunohistochemistry (phenotype)/FISH (genotype) approach to detect B-cells with rearrangements of BCR-ABL1 or 11q23/MLL post treatment in 44 and 7 patients, respectively. Cytospin slides, made from residual bone marrow, were stained with a monoclonal CD19 (clone HD37) antibody and scanned on an image analysis system (BioView Duet™) to target the CD19 positive B-cell population. During the scan, the location of each CD19 positive cell was recorded. The slides were subsequently destained and hybridized with FISH probes specific for the genotypic rearrangements mentioned above, with only antibody-targeted cells analyzed (target FISH or T-FISH). Disease was detected by T-FISH and at least one other method in 50 (39.7%) of the 126 samples tested, a finding comparable to the percentage identified by Q-RT-PCR (39.0%) and/or flow cytometry (23.2%). In samples with positive FISH results, T-FISH outperformed or was comparable to standard FISH in detecting disease in 47 (94%) samples. Importantly, T-FISH detected an abnormal cell population in 14/50 (28%) that standard FISH did not detect (p = 0.0064). Eight (57.1%) of these 14 samples had concurrent positive Q-RT-PCR results. The remaining six samples had MLL rearrangements and PCR studies were not performed. In three (6%) samples, abnormal cells were not CD19 positive and thus not detected until a followup area scan of the entire slide revealed low-level positivity in an apparent subset of CD19 negative progenitor B-cells. This latter finding was not observed in the concurrent negative controls. Only two samples (1.6%) with low level BCR-ABL1 positivity by Q-RT-PCR (>10−5) were negative by T-FISH. Serial dilution experiments of CD19-positive/t(9;22)-positive ALL-1 and CD19-negative/t(9;22)-negative Kasumi-1 cell lines demonstrated that T-FISH identified abnormalities at dilutions as low as 10−5, with consistent and reliable detection at 10−3. These observations suggest that antibody-targeted FISH is an effective way to increase the sensitivity of a slide-based assay in detecting residual “high risk” ALL.